1897] ORIGIN OF SPECIES AMONG PLANTS 169 
features may even be the actual extinction of a species; for it is 
conceivable that if a plant cannot set seed by self-fertilisation, and 
is not crossed by insects or the wind, it will die out, if it be an 
annual or not propagated by its vegetative system. It will thus be 
eliminated by natural selection.1 But the process falls within the 
subject of the distribution of species, both in time and space, and 
has nothing whatever to do with the origination of such harmful 
structures ; which, as long as they exist, are regarded as specific or 
generic characters. 
The survival of the fittest, therefore, and the destruction of the 
least fit and incapable to survive, are questions altogether inde- 
pendent of the Origination of Structural Variations, upon which 
the survival, or destruction in some cases, may actually depend. 
The reader must constantly bear in mind Darwin’s words which I 
again quote, because of their importance: “The direct action of the 
conditions of life . . . is a totally distinct consideration from the 
effects of natural selection . . . [it] has no relation whatever to the 
primary cause of any modification of structure.” ? This last is the 
sole matter with which I am concerned. 
Individual Differences.—These according to Darwin? and Dr 
Wallace are the chief materials for natural selection to act upon. 
As I have already fully discussed this subject in Matwral Science * 
and pointed out that as a rule they are quite incapable of giving rise 
to varietal characters which a systematist would take note of, I need 
say no more than invite the reader’s careful attention to my article. 
I might, however, briefly point out a fallacy in Dr Wallace’s 
conclusion. He has given numerous tables in his work, “ Darwinism,’ 
and argues that any excess in dimension of an organ from the mean 
is eliminated by natural selection ; so that a species keeps its dimen- 
sions pretty constant, annually.® But no intimation is given as to 
how great a deviation, in excess or deficiency of the mean, is required 
to prove destructive to the creature itself. Experience, however, 
shows that both nanism and gigantism are common phenomena in 
nature among plants; in which the customary deviations are vastly 
exceeded. Moreover they can be induced to arise under cultivation 
coupled with perfect health, fertility and heredity. Therefore, the 
whole of this argument falls to the ground. 
Supposed Requirements of Natural Selectioninthe For- 
mation of New Varieties.—The primary condition assumed by 
Darwin and Dr Wallace is a large population. In order to produce a 
new variety these writers tell us that “in the great majority of cases 
a new species arises amidst the population of an existing species.” 
Eee Omnoin, Cbe.,.- POs. 
2 Animals and Plants under Domestication,” vol. ii., p. 272. 
3“ Origin, etc.,” p. 34. 4 Vol. vi., p. 385 (1895). 
5 Journ. Linn. Soc. (Zool.), vol. xxv., p. 483. 
